728 ANTIGENIC PROPERTIES OF THE BACTERIAL CELL 



to as "specific soluble substances" or as "residue antigens,"' are carbohydrate in na- 

 ture and apparently free from protein, at least in significant amounts. They react 

 with antibodies, but fail to produce them, and hence may represent only a "haptene" 

 group (Landsteiner and Simms)^ of an unstable protein-gum complex existing in the 

 intact bacterial cell. They are probably of large molecular size since they do not pass 

 through dialyzing membranes. Their ultimate constitutents are the simple sugars or 

 sugar acids, but not always those best known. As Heidelberger has pointed out, biolog- 

 ical specificity, hitherto considered to be an attribute of protein structure alone, may 

 well be determined also by complex carbohydrates of this type. The considerable 

 number of simple sugars, hexoses as well as pentoses and perhaps others, together with 

 sugar acids and nitrogen-containing sugars — several asymmetric carbon atoms in each 

 — and a number of possible modes of linkage, make the possibilities of chemically dif- 

 ferent gum structure as wide as those in the protein group. 



BACTERIAL AUTOLYSATES 



In working with materials as vulnerable to chemical insult as are the biologically 

 active constituents of any cells, it is obvious that injury to antigenic function may 

 easily occur. It is necessary, therefore, to discuss briefly the antigenic properties of 

 bacterial autolysates which represent cell constituents in a state of suspension or solu- 

 tion produced by methods far less destructive than those employed in the manipula- 

 tions of extraction. 



Bacteria do not autolyze readily, and it is only with a few species, like the pneumo- 

 coccus, meningococcus, and some others, that the effects of autolysis can be stud- 

 ied at the present time. The remarks which follow, therefore, are based for the most 

 part on observations carried out with the pneumococcus. 



If one allows thick suspensions of this organism to autolyze, either in slightly 

 alkaline salt solution, in the presence of toluene or carbolic acid, or with the aid of 

 bile — which is probably an accelerator of autolysis'' — an antigen is produced which is 

 qualitatively identical, as to the antibodies produced, with the nucleoprotein fraction 

 obtained in the manner above recorded ; that is, the antibodies produced by such an 

 autolysate do not react in vitro with the specific carbohydrate substance, but react 

 with autolysate or nucleoprotein. It would seem, however, that though qualitatively 

 there is this resemblance, the autolysate represents the antigen in perhaps a more 

 potent, less denatured form, because of the more gentle method of its liberation. This 

 opinion is based upon the fact that such an autolysate, injected into guinea pigs, will 

 sensitize far more rapidly and vigorously than does similar treatment with extracted 

 nucleoprotein. 4 The skin reactions are severe, hemorrhagic, and eventually may ul- 

 cerate to a degree exceeding the most severe tuberculin reactions. With nucleoprotein 

 sensitization, on the other hand, nothing more definite than red, edematous lesions 



' While our original term for these substances was "residue antigen" and we here use the term 

 to avoid confusion, we believe that Dochez and Avery's nomenclature of "soluble specific substance" 

 is most accurately descriptive and should be adopted. 



2 Landsteiner, K., and Simms, S.: /. Exper. Med., 38, 127. 1923. 



3 Atkin, E. E.: Brit. J. Exper. Path., 7, 167. 1926. 

 '•Zinsser, H., and Grinnell, F. B.: J. Bad., 14, 301. 1927. 



